Gene knock-in using the CRISPR/Cas9 system can be achieved in a specific population of neurons in the mouse brain, by using in utero electroporation to introduce DNA fragments into neural progenitor cells. Using this strategy, we previously knocked-in the EGFP coding sequence into the N-terminal region of the beta-actin gene specifically in the pyramidal neurons in layer 2/3 of the somatosensory cortex. However, the knock-in efficiency was less than 2% of the transfected neurons. In this study, we sought to improve the knock-in efficiency using this system. First, we varied the length of the homology arms of the beta-actin donor template DNA, and found that the knock-in efficiency was increased to similar to 14% by extending the length of the 5' and 3' homology arms to 1.6 kb and 2.0 kb, respectively. We then tested the effect of the DNA repair protein RAD51 and the knock-in efficiency was increased up to 2.5-fold when co-transfecting with two different beta-actin and a camk2a targeting EGFP knock-in modules. The RAD51 overexpression did not alter the migration of developing neurons, density or morphology of the dendritic spines compared to those in neurons not transfected with RAD51. RAD51 expression will be useful for increasing the knock-in efficiency in neurons in vivo by CRISPR/Cas9-mediated homology directed repair (HDR). (C) 2020 Elsevier Inc. All rights reserved.